Data center liquid cooling method and system, storage medium and program product

Abstract

The invention discloses a data center liquid cooling method and system, a storage medium and a program product, and the method comprises the steps: judging whether a computer needs to change the operation state or not according to a to-be-completed task list; if yes, determining computer operation parameters corresponding to the tasks, and obtaining an operation parameter list; determining a cooling liquid parameter list corresponding to the operation parameter list in a preset table; determining a plurality of conveying pipeline lines; respectively determining weights of cooling liquid influence factors in the plurality of conveying pipeline lines; calculating an influence score according to the weight; calculating a conveying pipeline score list according to the cooling liquid parameters and the influence scores; respectively determining an optimal conveying pipeline score with the highest score in each conveying pipeline score list to obtain an optimal conveying pipeline score list; and generating an optimal conveying pipeline list according to the optimal conveying pipeline score list, and sending the optimal conveying pipeline list to the liquid cooling equipment. According to the computer cooling device, the computer is cooled in real time, and the instability of the computer is reduced.

Description

Technical Field

[0001] The present application belongs to the field of electronic digital data processing, and in particular, relates to a data center liquid cooling method, system, storage medium and program product. Background Art

[0002] Data centers are core facilities for modern computing and data storage. With the rapid development of technologies such as cloud computing, big data analysis, and artificial intelligence, the scale and computing needs of data centers are growing. This growth has brought huge energy consumption, especially in heat dissipation. The energy consumption of data centers has become an important factor in operating costs and environmental impact. Traditional data centers mostly use air cooling systems, but this method is inefficient and its efficiency is increasingly insufficient to cope with high heat loads as equipment density increases. Therefore, finding more efficient cooling methods has become an important need in the industry.

[0003] In the related art, key heat generation parameters are determined: identify and determine the key heat generation parameters of computing hardware (such as CPU, GPU) and memory hardware (such as DR data center liquid cooling method, system, storage medium and program product M) in the data center, such as hardware utilization and power. Obtain the water temperature, flow rate and computing hardware temperature at different values ​​of the key heat generation parameters of the computing hardware, and fit the computing hardware temperature function. Obtain the current temperature of the memory hardware, the water temperature and the temperature change rate of the key heat generation parameters at different values, and fit the memory hardware temperature change rate function. Calculate the water temperature and flow rate solution set when the hardware safety temperature is not exceeded according to the computing hardware temperature function. Calculate the water temperature solution set when the hardware safety temperature is not exceeded according to the memory hardware temperature change rate function. Select the optimal solution from the solution set for regulation. For computing hardware, adjust the water temperature and flow rate at the same time. For memory hardware, mainly adjust the water temperature, and set the flow rate to a fixed value.

[0004] However, the above control method only debugs the liquid cooling equipment of the data center at the current moment, but the working status of the computing hardware and memory hardware of the data center will change with the different running tasks, so the temperature and flow rate of the required coolant will also change. The debugging parameters obtained by applying the above technical solution may not be suitable for all work tasks, and there are also multiple options for the transportation route of the coolant. How to choose the transportation route with the least impact on the coolant is crucial to controlling the operating status of the computer. Summary of the invention

[0005] The present application provides a data center liquid cooling method, system, storage medium and program product, which are used to obtain the coolant parameters corresponding to each task by determining a list of tasks to be completed, cool the coolant parameters according to the characteristics of the delivery pipeline and formulate a transportation route that has the least impact on the coolant, and then enable the liquid cooling equipment to cool the computer according to the optimal transportation route and coolant parameters corresponding to the tasks to be completed, thereby achieving real-time cooling of the computer and reducing the instability of the computer.

[0006] In a first aspect, the present application provides a data center liquid cooling method, which determines whether a computer needs to change its operating state according to a list of tasks to be completed, wherein the list of tasks to be completed includes tasks to be completed after the current moment, and different operating states correspond to different tasks performed by the computer; If so, determine the computer operating parameters corresponding to each task according to the order of tasks in the to-be-completed task list, and obtain an operating parameter list in which the arrangement order corresponds to the task order; Determine a coolant parameter list corresponding to the operating parameter list in a preset table, the operating parameters correspond to the coolant parameters one by one, and the coolant parameters include coolant temperature and coolant flow rate; Determine several delivery pipeline routes for the liquid cooling equipment to deliver cooling liquid to the computer; Determine the weights of coolant influencing factors in several pipeline routes respectively, the coolant influencing factors include the number of computers passing through, pipeline material and pipeline length; The impact score corresponding to each transmission pipeline line is calculated based on the weight of the coolant impact factor; A pipeline score list is calculated based on the coolant parameters and the impact score of each pipeline line. The pipeline score list includes each task to be completed. Each task to be completed includes a corresponding pipeline score. Each task to be completed and the corresponding pipeline score constitute a pipeline score sublist. Each pipeline score sublist constitutes a pipeline score list. Determine the best pipeline score with the highest score in each pipeline score sublist respectively, and obtain the best pipeline score list, wherein the best pipeline score corresponds to the best pipeline route; An optimal delivery pipeline list is generated according to the optimal delivery pipeline score list, and the optimal delivery pipeline list is sent to the liquid cooling device. The arrangement order of the optimal delivery pipeline list corresponds to the task order of the to-be-completed task list.

[0007] By adopting the above technical solution, first determine whether the computer needs to change its operating state according to the task list. If necessary, determine the operating parameters corresponding to each task, and find the matching coolant temperature and flow rate parameters in the preset table. Then, the method comprehensively considers multiple influencing factors of the delivery pipeline, including the number of computers passing through, the thermal conductivity of the pipe, and the length of the pipeline, and calculates the impact score of each pipeline by assigning different weights. On this basis, combined with the coolant parameters and pipeline impact scores, a delivery pipeline score list is obtained, and the optimal delivery scheme under each task is screened out. Finally, the optimized delivery pipeline list is sent to the liquid cooling equipment for execution. According to the real-time task requirements of the data center, the computer operating parameters and cooling system are dynamically adjusted to avoid the problems of over-cooling or insufficient cooling in the traditional scheme, and reduce energy waste. At the same time, the differential impact of the coolant delivery pipeline is considered. By introducing the weight of the influencing factor and the scoring mechanism, the optimal delivery scheme under various working conditions is found, which further improves the efficiency and accuracy of the liquid cooling system.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, determining a plurality of delivery pipelines through which a liquid cooling device delivers cooling liquid to a computer specifically includes: Obtain a piping layout diagram for connecting the liquid cooling device to the computer via a delivery piping line; The pipeline layout diagram is calculated using a path recognition algorithm to obtain a number of pipeline routes for the liquid cooling device to transport the cooling liquid to the computer; According to the pipeline layout diagram, the number of path computers, pipeline materials and pipeline lengths corresponding to several transmission pipeline lines are determined respectively. The number of path computers, pipeline materials and pipeline lengths are coolant influencing factors.

[0009] By adopting the above technical solution, it is necessary to first obtain the pipeline layout diagram, and then use the path recognition algorithm to automatically calculate all possible transmission routes. This method can quickly and accurately find the effective path in the complex pipeline network inside the data center, and extract the key parameters affecting the cooling effect according to the pipeline layout diagram, including the number of computers passing through the pipeline, the properties of the pipe, and the length of the pipe. These three parameters reflect the characteristics of the transmission pipeline from different angles: the number of computers passing through affects the distribution of cooling load, the properties of the pipe determine the heat conduction efficiency, and the pipe length is related to the fluid resistance and transmission loss. By collecting these parameters, the subsequent evaluation algorithm can analyze the advantages and disadvantages of each pipeline route more comprehensively and finely, providing a reliable basis for finding the best transmission solution. The automatic search and parameter extraction of the transmission line are realized, and the control efficiency and accuracy of the liquid cooling system are improved.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, weights of coolant influencing factors in a plurality of delivery pipeline lines are determined respectively, specifically including: Determine whether the number of path computers is greater than a preset number; If it is greater than the preset number, the weight of the number of computers passed is set to the first weight; If it is not greater than the preset number, the weight of the number of computers in the path is set to the second weight; Determine the average thermal conductivity of the pipe material; Determine whether the average thermal conductivity is greater than a preset thermal conductivity; If it is greater than the preset thermal conductivity, the weight of the pipe material is set to the third weight; If it is not greater than the preset thermal conductivity, the weight of the pipe material is set to the fourth weight; The weight of the pipeline length is set to the fifth weight, and the sum of the weight of the number of computers passing through, the weight of the pipeline material and the weight of the pipeline length is equal to the preset threshold.

[0011] By adopting the above technical solution, the method introduces a preset threshold for the number of computers. When the actual number exceeds the threshold, a larger weight is given, otherwise a smaller weight is given. This way of distinguishing weights can highlight the impact of pipeline loads, give more weights to pipelines in computer-intensive areas, and ensure the reasonable allocation of cooling resources. At the same time, a similar threshold comparison strategy is also adopted for thermal conductivity. Different weights are given according to the actual thermal conductivity of the pipe to emphasize the impact of material selection on heat transfer efficiency. In addition, the method also stipulates that the sum of the weights of the three influencing factors must be equal to the preset value, forming a dynamic balance weight allocation mechanism. This means that when the weight of a factor increases, the weights of other factors will decrease accordingly, thereby maintaining the total weight unchanged. On the one hand, this weight allocation method ensures the stability of the impact assessment and avoids the excessive dominance of a certain factor. On the other hand, it also increases the flexibility of weight assignment, and can adjust the emphasis of different factors according to actual needs.

[0012] In combination with some embodiments of the first aspect, in some embodiments, the influence score corresponding to each delivery pipeline line is obtained according to the weight calculation of the coolant influence factor, specifically including: Determine the parameter value of the coolant influencing factor corresponding to each delivery pipeline line; The parameter values ​​corresponding to each transmission pipeline route and the weight of the coolant influencing factor are respectively input into the influence scoring function to obtain the influence score corresponding to each transmission pipeline route.

[0013] By adopting the above technical solution, it is necessary to first obtain the parameter values ​​of the influencing factors corresponding to each pipeline, and then input the parameter values ​​and weights into the influence scoring function, and the influence score of the pipeline is calculated by the function. The influence scoring function uniformly expresses the effects of multiple influencing factors in a quantitative way, providing an operational mathematical model for the comprehensive evaluation of pipeline performance. Through function calculation, the degree of influence of different factors is converted into a unified scoring value, making the pipeline properties that were originally difficult to directly compare measurable and sortable. This quantitative evaluation method overcomes the subjectivity and ambiguity of qualitative analysis and enhances the scientificity and accuracy of pipeline selection. On the other hand, the design of the influence scoring function also provides an important basis for the optimization and control of the liquid cooling system. By analyzing the scoring results, the pros and cons of each pipeline can be clearly judged and the key factors affecting the heat transfer efficiency can be identified. This information can guide engineers to make targeted improvements in pipeline layout, material selection and other aspects, and continuously improve the performance of the liquid cooling system. At the same time, the pipeline impact score also lays the foundation for subsequent advanced strategies such as flow rate and temperature optimization and multi-pipeline collaborative control, enabling the liquid cooling system to achieve intelligent and refined regulation in a larger range.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, the influence scoring function is: In the formula, To score the impact, is the number of computers in the path, is the first weight or the second weight, is the average thermal conductivity, is the third weight or the fourth weight, is the pipe length, The fifth weight.

[0015] By adopting the above technical solution, the function comprehensively considers the three influencing factors of the number of computers in the passage, the thermal conductivity of the pipe and the length of the pipeline, and uniformly expresses their effects as a score value by weighted summation. This quantitative evaluation method overcomes the subjectivity and uncertainty of traditional qualitative analysis, and provides a reliable mathematical basis for the comparison and optimization of pipeline performance. The design of the scoring function is reasonable, concise and has a clear physical meaning. Among them, the number of computers reflects the cooling load borne by the pipeline, the thermal conductivity represents the heat transfer performance of the pipe, and the pipe length affects the resistance loss of the fluid. By introducing the weight coefficient, the function can flexibly adjust the importance of each factor to meet the needs of different application scenarios. At the same time, the sum of the weights in the function is limited to 1, which is both a normalization process and ensures the stability and comparability of the scoring results.

[0016] In combination with some embodiments of the first aspect, in some embodiments, a pipeline score list is obtained by calculating the impact score of the coolant parameters and each pipeline line, specifically including: Inputting the coolant parameters and the impact score of each pipeline route into a pipeline score calculation function to obtain a pipeline score of each pipeline route; Add the delivery pipeline scores corresponding to the coolant parameters to the corresponding blank list to obtain a delivery pipeline score list; The pipeline score calculation function is: In the formula, Score the pipeline. is the coolant temperature, is the coolant flow rate, Score impact; exist and When the coolant temperature is fixed, the higher the coolant temperature, the smaller the effect of the delivery pipeline on the coolant. The bigger; exist and When the coolant flow rate is a fixed value, the greater the coolant flow rate, the greater the impact of the delivery pipeline on the coolant. The smaller; exist and When the impact score is a fixed value, the greater the impact of the delivery pipeline on the coolant, the greater the impact of the delivery pipeline on the coolant. The smaller.

[0017] By adopting the above technical scheme, the score integrates the effects of factors such as coolant temperature, flow rate and pipeline impact score, and quantifies the comprehensive performance of the pipeline through a mathematical function. The pipeline score reflects more the mutual influence between the pipeline and the coolant, and reflects the overall effect of the liquid cooling system. In the design of the pipeline scoring function, the temperature, flow rate and impact score appear in the form of exponential function, reciprocal and reciprocal respectively. The selection of this form has a clear physical meaning. First, the exponential function is a common attenuation model. As the temperature increases, its reciprocal decays exponentially, indicating the nonlinear relationship between the negative impact of temperature on the cooling effect and the temperature increase. Secondly, the flow rate and the reciprocal of the impact score reflect the inverse relationship between them and the heat transfer performance of the pipeline, that is, the greater the flow rate and the higher the impact score, the worse the heat transfer effect per unit length of the pipeline. By comprehensively considering the effects of temperature, flow rate and impact score, and using the characteristics of reciprocal operation, the pipeline scoring function cleverly unifies their influences into a dimensionless score value.

[0018] In combination with some embodiments of the first aspect, in some embodiments, after generating an optimal delivery pipeline list according to the optimal delivery pipeline score list and sending the optimal delivery pipeline list to the liquid cooling device, the method further includes: If not, the coolant parameters and delivery pipeline at the current moment are maintained to cool the computer.

[0019] By adopting the above technical solution, when the system determines that there is no need to change the operating state, the current coolant parameters and delivery pipeline selection are maintained. Once it is determined that the existing cooling configuration can meet the task requirements, the system will maintain the current working parameters and avoid unnecessary adjustment behavior.

[0020] In a second aspect, an embodiment of the present application provides a data center liquid cooling system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and one or more processors call the computer instructions to enable the data center liquid cooling system to execute the method described in the first aspect and any possible implementation method of the first aspect.

[0021] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, comprising instructions. When the instructions are executed on a data center liquid cooling system, the data center liquid cooling system executes the method described in the first aspect and any possible implementation method of the first aspect.

[0022] In a fourth aspect, an embodiment of the present application provides a computer program product, characterized in that when the computer program product runs on a data center liquid cooling system, the data center liquid cooling system executes the method described in any possible implementation manner in the first aspect.

[0023] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. The present application provides a data center liquid cooling method. First, it is determined whether the computer needs to change the operating state according to the task list. If necessary, the operating parameters corresponding to each task are determined, and the matching coolant temperature and flow rate parameters are found in the preset table. Then, the method comprehensively considers multiple influencing factors of the delivery pipeline, including the number of computers passing through, the thermal conductivity of the pipe, and the length of the pipeline, and calculates the impact score of each pipeline by assigning different weights. On this basis, combined with the coolant parameters and the pipeline impact score, a delivery pipeline score list is obtained, and the optimal delivery scheme under each task is screened out. Finally, the optimized delivery pipeline list is sent to the liquid cooling equipment for execution. According to the real-time task requirements of the data center, the computer operating parameters and cooling system are dynamically adjusted to avoid the problems of over-cooling or insufficient cooling in the traditional scheme, and reduce energy waste. At the same time, the differential impact of the coolant delivery pipeline is taken into account. By introducing the weight of the influencing factor and the scoring mechanism, the optimal delivery scheme under various working conditions is found, which further improves the efficiency and accuracy of the liquid cooling system.

[0024] 2. The present application provides a data center liquid cooling method, which first requires obtaining a pipeline layout diagram, and then using a path recognition algorithm to automatically calculate all possible transmission routes. This method can quickly and accurately find the effective path in the complex pipeline network inside the data center, and extract the key parameters that affect the cooling effect according to the pipeline layout diagram, including the number of computers passing through the pipeline, the properties of the pipe, and the length of the pipe. These three parameters reflect the characteristics of the transmission pipeline from different angles: the number of computers passing through affects the cooling load distribution, the properties of the pipe determine the heat conduction efficiency, and the pipe length is related to the fluid resistance and transmission loss. By collecting these parameters, the subsequent evaluation algorithm can analyze the advantages and disadvantages of each pipeline route more comprehensively and finely, providing a reliable basis for finding the best transmission solution. The automatic search and parameter extraction of the transmission line are realized, and the control efficiency and accuracy of the liquid cooling system are improved.

[0025] 3. The present application provides a data center liquid cooling method. First, it is necessary to obtain the influencing factor parameter values ​​corresponding to each pipeline, and then input the parameter values ​​and weights into the impact scoring function, and the impact score of the pipeline is calculated by the function. The impact scoring function uniformly expresses the effects of multiple influencing factors in a quantitative way, providing an operational mathematical model for the comprehensive evaluation of pipeline performance. Through function calculation, the degree of influence of different factors is converted into a unified scoring value, so that the pipeline properties that were originally difficult to directly compare have measurable and sortable characteristics. This quantitative evaluation method overcomes the subjectivity and ambiguity of qualitative analysis and enhances the scientificity and accuracy of pipeline selection. On the other hand, the design of the impact scoring function also provides an important basis for the optimization and control of the liquid cooling system. By analyzing the scoring results, the pros and cons of each pipeline can be clearly judged and the key factors affecting the heat transfer efficiency can be identified. This information can guide engineers to make targeted improvements in pipeline layout, material selection and other aspects, and continuously improve the performance of the liquid cooling system. At the same time, the pipeline impact score also lays the foundation for subsequent advanced strategies such as flow rate temperature optimization and multi-pipeline collaborative control, so that the liquid cooling system can achieve intelligent and refined regulation in a larger range. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of a data center liquid cooling method in an embodiment of the present application.

[0027] Figure 2 This is another flow chart of a data center liquid cooling method in an embodiment of the present application.

[0028] Figure 3 It is a schematic diagram of the physical device structure of a data center liquid cooling system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0029] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification and appended claims of the present application, the singular expressions "one", "a kind of", "said", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to any or all possible combinations comprising one or more listed items.

[0030] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more.

[0031] An application scenario of an embodiment of the present application is described below: In a large Internet company data center, server cabinets are neatly arranged and high-performance computing equipment is running non-stop. These devices carry the company's key business systems, including user data storage, real-time transaction processing, data mining and analysis, etc. In order to ensure the efficient operation of the business, the data center management team needs to monitor the operating status of the equipment at all times, especially the temperature changes of the equipment.

[0032] Recently, the data center has launched a batch of high-density computing clusters to handle the company's newly launched artificial intelligence business. These powerful machine learning algorithms consume a lot of computing resources, and the heat generation of the equipment has also increased significantly. The traditional cooling system has been unable to meet the needs of rapid heat dissipation. The temperature of the equipment frequently exceeds the safety threshold, and there have even been several restart accidents caused by overheating, affecting business continuity. The data center management team realized that the cooling system needed to be upgraded as soon as possible to meet the heat dissipation needs of the equipment in the new business scenario. After repeated evaluations, they decided to introduce an advanced liquid cooling system to remove the heat generated by the equipment through the circulation of coolant. However, how to adjust the temperature and flow rate of the coolant according to the real-time working conditions of the equipment to ensure the heat dissipation effect without wasting energy is a key optimization problem. The operation and maintenance engineer staff was responsible for the implementation of this project. First, he analyzed the key heat generation parameters of different types of equipment and established the functional relationship between the equipment temperature and the coolant temperature and flow rate. On this basis, he developed a liquid cooling control system to collect the working status data of the equipment in real time, dynamically calculate the optimal coolant parameters, and send them to the actuator for adjustment.

[0033] However, the staff found that the cooling pipelines inside the data center are complicated and there are multiple optional coolant transportation paths. The pipe lengths and fluid resistances of different paths are different, and the dynamic adjustment of coolant parameters will have different effects on different pipelines. For example, some pipelines may not be able to withstand the rapid changes in temperature and be damaged; some pipelines far away from the heat source may have cooling lag problems, making it difficult to fully match the real-time cooling needs of the equipment.

[0034] In addition, due to the scheduling strategy of the business system, the type and intensity of computing tasks carried by the same device in different time periods may also change, thereby affecting the heat dissipation requirements. For complex machine learning jobs, the previous second may be a high-intensity matrix operation, and the next second may turn into relatively idle data reading and writing. How to predict the heat dissipation trend of the equipment, adjust the coolant parameters in advance, and avoid drastic fluctuations in the temperature curve is also a major challenge. In order to solve the above technical problems, the present application provides a data center liquid cooling method for obtaining the coolant parameters corresponding to each task by determining a list of tasks to be completed, cooling the coolant parameters according to the characteristics of the transport pipeline, and formulating a transportation route that has the least impact on the coolant, and then using the liquid cooling equipment to cool the computer according to the optimal transportation route and coolant parameters corresponding to the task to be completed, thereby achieving real-time cooling of the computer and reducing the instability of the computer.

[0035] Combine the following Figure 1 , a data center liquid cooling method in an embodiment of the present application is described: See also Figure 1 , which is a flow chart of a data center liquid cooling method in an embodiment of the present application.

[0036] S101, judging whether the computer needs to change its running state according to the list of tasks to be completed; The system determines whether the computer needs to change its operating state based on the list of tasks to be completed. The list of tasks to be completed contains tasks to be completed after the current moment. The computer has different operating states when performing different tasks. In this step, the system first needs to obtain the list of tasks to be completed, which contains all the tasks that the computer needs to perform after the current moment. Different tasks may have different requirements for the computer's operating state. For example, some tasks require high-frequency CPU operations, while other tasks may rely mainly on memory operations. The system determines whether the computer needs to adjust its operating state accordingly to meet the requirements of the task by analyzing the list of tasks to be completed.

[0037] For example, suppose the current time is t0, and the list of tasks to be completed is [Task A, Task B, Task C]. Task A needs to be executed at t1, which has high requirements for the CPU frequency; Task B needs to be executed at t2, which has high requirements for memory bandwidth; Task C needs to be executed at t3, which has relatively low requirements for CPU and memory. By analyzing this task list, the system finds that at t1 and t2, the computer's operating state needs to be adjusted to meet the performance requirements of Task A and Task B. At t3, the computer can maintain a lower operating state to save energy.

[0038] In terms of technical implementation, the system can use task scheduling algorithms to analyze the list of tasks to be completed, and determine the optimal operating state of the computer at different times based on the resource requirements and time constraints of each task. Common task scheduling algorithms include but are not limited to: First Come First Served (FCFS) algorithm, Shortest Job First (SJF) algorithm, time slice round-robin algorithm, priority scheduling algorithm, etc. The system can select a suitable scheduling algorithm based on actual conditions to take into account both the response time of the task and the utilization efficiency of computing resources.

[0039] S102, determining the computer operating parameters corresponding to each task according to the order of tasks in the to-be-completed task list, and obtaining an operating parameter list in which the arrangement order corresponds to the order of tasks; If so, the system determines the computer operating parameters corresponding to each task according to the order of tasks in the to-be-completed task list, and obtains an operating parameter list in which the arrangement order corresponds to the task order one by one.

[0040] After determining that the computer needs to change its operating state, the system needs to further determine the specific requirements of each task for the computer's operating parameters. The operating parameters may include but are not limited to: CPU main frequency, memory frequency, power consumption budget, etc. The system analyzes the resource requirements of each task one by one according to the order of the tasks in the list of tasks to be completed, and thus obtains a list of operating parameters that corresponds to the order of the tasks.

[0041] For example, for the task list [Task A, Task B, Task C] in the previous step, the system may obtain the following list of operating parameters: (CPU main frequency = 3.0GHz, memory frequency = 2400MHz, power consumption budget = 120W) (CPU frequency = 2.5GHz, memory frequency = 3200MHz, power budget = 110W) (CPU main frequency = 1.8GHz, memory frequency = 2133MHz, power consumption budget = 70W) Among them, the first group of operating parameters corresponds to task A, the second group corresponds to task B, and the third group corresponds to task C. In this way, the system establishes a mapping relationship between tasks and operating parameters.

[0042] In terms of technical implementation, the system can use a prediction model to estimate the resource requirements of each task. The prediction model can be trained based on historical data through machine learning algorithms (such as regression analysis, neural networks, etc.). When a new task arrives, the system inputs the characteristic information of the task (such as task type, input data volume, etc.) into the prediction model. The model gives the performance of the task under different operating parameters and selects the best parameter combination that meets the task requirements. At the same time, the system also needs to consider the physical limitations of the computer, such as maximum power consumption, temperature thresholds, etc., to ensure that the operating parameters are set within a safe range.

[0043] S103, determining a coolant parameter list corresponding to the operating parameter list in a preset table; The system determines a coolant parameter list corresponding to the operating parameter list in a preset table, and the operating parameters correspond one-to-one to the coolant parameters. The coolant parameters include the coolant temperature and the coolant flow rate.

[0044] After obtaining the computer's operating parameter list, the system needs to set the appropriate coolant parameters according to these parameters to ensure that the computer can be effectively cooled when running under high load. The coolant parameters mainly include the temperature and flow rate of the coolant. The system can pre-establish a mapping table to associate different operating parameter combinations with the optimal coolant parameters. When a new operating parameter list is generated, the system only needs to query in the table to quickly determine the corresponding coolant parameter list.

[0045] For the operating parameter list obtained in the previous step, the system matches them one by one in the table to obtain the corresponding coolant parameter list: (Temperature = 20°C, Flow rate = 5L / min) (Temperature = 22°C, Flow rate = 4L / min) (Temperature = 25°C, Flow rate = 2L / min) In this way, the system establishes a corresponding relationship between operating parameters and coolant parameters.

[0046] In terms of technical implementation, the preset table can be stored in a key-value pair data structure for fast query operations. At the same time, the system can also introduce certain fault-tolerant mechanisms. For example, when the operating parameters do not completely match the data in the table, the most similar parameter combination is used, or the adapted coolant parameters are obtained through interpolation calculation. In addition, the preset table needs to be updated regularly to adapt to possible hardware changes or changes in cooling requirements.

[0047] S104, determining a number of delivery pipelines for the liquid cooling device to deliver the cooling liquid to the computer; After determining the temperature and flow rate parameters of the coolant, the system needs to further plan the delivery path of the coolant. Generally speaking, there is not only one coolant delivery pipeline from the liquid cooling equipment to the computer, but there are several optional routes. Different routes may differ in physical properties such as pipeline length, number of turns, and material, which in turn affects the transmission effect and efficiency of the coolant. Therefore, the system needs to comprehensively evaluate each delivery pipeline route to provide a basis for subsequent path optimization. The specific implementation of this step will be described in detail below and will not be repeated here.

[0048] S105, respectively determining the weights of coolant influencing factors in a number of delivery pipelines; The system determines the weights of coolant influencing factors in several pipeline routes respectively. The coolant influencing factors include the number of computers passing through, the material of the pipeline and the length of the pipeline. Specifically: determine whether the number of computers passing through is greater than the preset number; If it is greater than the preset number, the weight of the number of computers passed is set to the first weight; If it is not greater than the preset number, the weight of the number of computers in the path is set to the second weight; Determine the average thermal conductivity of the pipe material; Determine whether the average thermal conductivity is greater than a preset thermal conductivity; If it is greater than the preset thermal conductivity, the weight of the pipe material is set to the third weight; If it is not greater than the preset thermal conductivity, the weight of the pipe material is set to the fourth weight; The weight of the pipeline length is set to the fifth weight, and the sum of the weight of the number of computers passing through, the weight of the pipeline material and the weight of the pipeline length is equal to the preset threshold.

[0049] After identifying several optional coolant delivery pipeline routes, the system needs to further evaluate the pros and cons of each route. Here, the coolant impact factor is introduced to measure the impact of the physical characteristics of the pipeline route on the coolant transmission effect. The main influencing factors considered include: the number of computers passing through, the material of the pipeline, the length of the pipeline, etc. The system needs to set corresponding weights for these influencing factors so as to evaluate their importance in the subsequent comprehensive scoring.

[0050] For example, for the three pipelines mentioned in the previous step, the system can adopt the following weight setting strategy: For the number of computers: If there are more than 5 units, the weight is 0.4 (the first weight); If the number is not more than 5, the weight is 0.3 (second weight).

[0051] For pipe material: According to the thermal conductivity of various common materials, a threshold is set, for example, 200W / (m·K); If the average thermal conductivity of the line is greater than the threshold, the weight is 0.35 (third weight); If the average thermal conductivity of the line is not greater than the threshold, the weight is 0.25 (the fourth weight).

[0052] For pipe length: Taking into account that the effect of length on the liquid cooling system is relatively small, the weight is uniformly set to 0.3 (the fifth weight).

[0053] It should be noted that the sum of the weights of these three types of influencing factors should be 1 to ensure that there will be no deviation in the subsequent weighted calculation process.

[0054] In the above example, line A passes through 4 computers, and the average thermal conductivity is 16W / (m·K). Therefore, the weight of the number of computers passed is 0.3, the weight of the pipe material is 0.25, and the weight of the pipe length is 0.3.

[0055] Line B passes through 2 computers, and its average thermal conductivity is 401 W / (m·K). Therefore, the weight of the number of computers passed is 0.3, the weight of the pipe material is 0.35, and the weight of the pipe length is 0.3.

[0056] Line C passes through 5 computers, and the average thermal conductivity is 237W / (m·K). Therefore, the weight of the number of computers passed is 0.4, the weight of the pipe material is 0.35, and the weight of the pipe length is 0.3.

[0057] In terms of technical implementation, the weights of influencing factors can be set using a heuristic approach and adjusted based on experience and test data. At the same time, the system can also introduce an adaptive mechanism to dynamically optimize the weight parameters based on historical operating data and feedback information to adapt to actual application scenarios. In addition, if there is a clear correlation between influencing factors, it may be necessary to introduce more complex mathematical models, such as principal component analysis (PCA), to eliminate multicollinearity between variables and improve the accuracy of the comprehensive score.

[0058] S106. Calculate the impact score corresponding to each transmission pipeline line according to the weight of the coolant impact factor; The system calculates the impact score corresponding to each delivery pipeline line according to the weight of the coolant impact factor. Specifically: determine the parameter value of the coolant impact factor corresponding to each delivery pipeline line; The parameter values ​​corresponding to each transmission pipeline route and the weight of the coolant influencing factor are respectively input into the influence scoring function to obtain the influence score corresponding to each transmission pipeline route.

[0059] The influence scoring function is: In the formula, To score the impact, is the number of computers in the path, is the first weight or the second weight, is the average thermal conductivity, is the third weight or the fourth weight, is the pipe length, The fifth weight.

[0060] After determining the weight of the coolant influencing factor, the system needs to further calculate the comprehensive impact score of each pipeline route. The purpose of this step is to quantitatively evaluate the pros and cons of each route and provide a basis for subsequent route selection. The impact score is calculated using the weighted average method, multiplying the quantified value of each influencing factor by the corresponding weight, and then summing them up to obtain the final score result.

[0061] For the three pipeline lines mentioned in the previous step, the system first calculates the quantitative value of each influencing factor based on the physical parameters of the lines: Route A: Number of computers in the pipeline: 4; Thermal conductivity of pipeline material: 16W / (m·K); Length of pipeline: 20m.

[0062] Route B: Number of computers in the pipeline: 2; Thermal conductivity of pipeline material: 401W / (m·K); Length of pipeline: 25m.

[0063] Route C: Number of computers in the pipeline: 5; Thermal conductivity of pipeline material: 237W / (m·K); Length of pipeline: 18m.

[0064] The system then substitutes these quantitative values ​​into the influence scoring function for calculation.

[0065] S = N × m1 + K × m2 + H × m3 Among them, S is the impact score, N is the number of computers in the route, m1 is the first weight or the second weight, K is the average thermal conductivity, m2 is the third weight or the fourth weight, H is the pipeline length, and m3 is the fifth weight. Substituting the parameters of each line, we can get: Impact score of route A: S_A = 4 × 0.3 + 16 × 0.4 + 20 × 0.3 = 12.6 Impact score for route B: S_B = 2 × 0.3 + 401 × 0.4 + 25 × 0.3 = 168.5 Impact score of line C: S_C = 5 × 0.4 + 237 × 0.3 + 18 × 0.3 = 78.5 In this way, the system obtains the impact score of each of the three lines. The higher the score, the better the overall transmission performance of the line and the smaller the impact on the coolant.

[0066] S107, calculating and obtaining a pipeline scoring list according to the coolant parameters and the impact score of each pipeline line; The system calculates the pipeline score list based on the coolant parameters and the impact score of each pipeline route. The pipeline score list contains each task to be completed, each task to be completed contains the corresponding pipeline score, each task to be completed and the corresponding pipeline score constitute a pipeline score sublist, and each pipeline score sublist constitutes a pipeline score list. Specifically: the coolant parameters and the impact score of each pipeline route are input into the pipeline score calculation function to obtain the pipeline score of each pipeline route; Add the delivery pipeline scores corresponding to the coolant parameters to the corresponding blank list to obtain a delivery pipeline score list; The pipeline score calculation function is: In the formula, Score the pipeline. is the coolant temperature, is the coolant flow rate, Score impact; exist and When the coolant temperature is fixed, the higher the coolant temperature, the smaller the effect of the delivery pipeline on the coolant. The bigger; exist and When the coolant flow rate is a fixed value, the greater the coolant flow rate, the greater the impact of the delivery pipeline on the coolant. The smaller; exist and When the impact score is a fixed value, the greater the impact of the delivery pipeline on the coolant, the greater the impact of the delivery pipeline on the coolant. The smaller.

[0067] After obtaining the impact score of each pipeline line, the system needs to further combine the coolant temperature and flow rate parameters to calculate the final comprehensive score, namely the pipeline score. Different from the impact score, the pipeline score not only considers the characteristics of the pipeline line itself, but also the state parameters of the coolant, so it can more comprehensively reflect the transmission performance and efficiency of the pipeline line.

[0068] The higher the temperature, the higher the pipe score. This is because, with other conditions unchanged, the higher the temperature means the coolant carries more heat and the higher the transfer efficiency. The higher the flow rate, the lower the pipe score. This is because too high a flow rate will increase turbulence and pressure loss in the pipe, which is not conducive to heat transfer. The higher the impact score, the lower the pipe score. The impact score reflects the transmission performance of the pipeline line itself. The higher the score, the better the characteristics of the pipe and the less impact on the coolant.

[0069] For example, suppose the coolant parameters corresponding to the three pipelines to be evaluated are as follows: Line A: temperature is 20°C, flow rate is 5L / min Line B: temperature 22°C, flow rate 4L / min Line C: temperature 25°C, flow rate 2L / min Combined with the impact score calculated in the previous step, and substituted into the calculation formula of the transmission pipeline score, we can get: Pipeline score of line A: 3.28×10^(-8) Pipeline score of line B: 1.53×10^(-9) Pipeline score of line C: 6.87×10^(-9) Since there is more than one pipeline route to be evaluated, the system needs to generate a pipeline scoring list. Each element of this list corresponds to the scoring result of a pipeline route, and the order of the elements is consistent with the number of the pipeline route. In the above example, the generated pipeline scoring list is: [3.28 × 10^(-8), 1.53 × 10^(-9), 6.87 × 10^(-9)] This list reflects the comprehensive transmission performance of the three pipeline routes under the current coolant parameters, providing a quantitative basis for subsequent route selection and scheduling.

[0070] S108, respectively determining the optimal transport pipeline score with the highest score in each transport pipeline score sublist to obtain an optimal transport pipeline score list; The system determines the best transmission pipeline score with the highest score in each transmission pipeline score sublist, and obtains the best transmission pipeline score list, wherein the best transmission pipeline score corresponds to the best transmission pipeline route.

[0071] After obtaining the pipeline score list, the system needs to select the pipeline route with the highest score for each task to be completed as the actual coolant delivery channel. The purpose of this step is to match the optimal transmission path for each task based on a comprehensive evaluation of pipeline performance and coolant properties, thereby maximizing the efficiency and effectiveness of the liquid cooling system.

[0072] Specifically, the system first splits the pipeline scoring list according to the task dimension to obtain several pipeline scoring sublists, each of which corresponds to a task to be completed. Taking the scoring list in the previous step as an example, the split results are as follows: Pipeline score sublist for task A: [0.0164, 0.0012, 0.0020] Task B pipeline score sublist: [0.0382, 0.0029, 0.0047] Pipeline score sublist for task C: [0.0651, 0.0049, 0.0080] Then, the system compares the scores in each sublist, selects the subscript position with the maximum value, and records it. The subscript position here corresponds to the pipeline route (for example, subscript 0 represents route A, subscript 1 represents route B, and subscript 2 represents route C).

[0073] For Task A, the maximum score is 0.0164, corresponding to subscript 0, so the optimal delivery pipeline for Task A is Route A; For Task B, the maximum score is 0.0382, corresponding to subscript 0, so the optimal delivery pipeline for Task B is also Route A; For task C, the maximum score is 0.0651, corresponding to subscript 0, so the optimal delivery pipeline for task C is also route A.

[0074] Finally, the system organizes the scores of these optimal pipelines into a new list according to the task order, namely the optimal pipeline score list: [0.0164, 0.0382, 0.0651] This list directly gives the expected score of each task to be completed under the optimal delivery pipeline route, providing a decision-making basis for subsequent coolant scheduling and flow control.

[0075] S109, generating an optimal delivery pipeline list according to the optimal delivery pipeline score list, and sending the optimal delivery pipeline list to the liquid cooling device; The system generates an optimal delivery pipeline list according to the optimal delivery pipeline score list, and sends the optimal delivery pipeline list to the liquid cooling device. The arrangement order of the optimal delivery pipeline list corresponds to the task order of the task list to be completed.

[0076] After generating the optimal pipeline scoring list, the system needs to further convert the scoring results into specific pipeline route selection instructions and send them to the liquid cooling equipment to guide the actual flow and scheduling of the coolant. The purpose of this step is to implement the previous calculation results to the physical level and directly control and optimize the operation process of the liquid cooling system.

[0077] Specifically, the system first generates an optimal pipeline list corresponding to the task order based on the optimal route subscripts of each task in the optimal pipeline scoring list. Taking the scoring list obtained in the previous step as an example, the generation process of the optimal pipeline list is as follows: Optimal pipeline score list: [0.0164, 0.0382, 0.0651] The corresponding optimal line index: [0, 0, 0] The final optimal pipeline list is: [Line A, Line A, Line A] This list clearly gives the optimal coolant delivery pipeline corresponding to each task to be completed. The liquid cooling equipment can directly set the switch status of the valve according to this list to control the flow direction and distribution ratio of the coolant between different pipeline lines.

[0078] Then, the system sends the optimal pipeline list to the liquid cooling device through the communication interface. Here, the list information needs to be packaged into a data frame or message queue in a specific format according to the control protocol and interface specifications of the device. Common communication methods include but are not limited to: RS232 / RS485 serial communication, TCP / IP network communication, Modbus bus communication, etc.

[0079] S110, maintaining the current coolant parameters and the delivery pipeline to cool the computer.

[0080] If not, the coolant parameters and delivery pipeline at the current moment are maintained to cool the computer.

[0081] After sending the optimal pipeline list to the liquid cooling device, if the result is that the current computer operation status does not need to be changed, the system will maintain the existing coolant parameter settings and pipeline selection and continue to cool the computer. The purpose of this step is to maintain the stable operation of the liquid cooling system when the load demand does not change, avoid frequent parameter adjustments and pipeline switching, thereby reducing energy consumption and wear, and improving the reliability and service life of the system.

[0082] The above embodiments have the following beneficial effects: First, it is determined whether the computer needs to change its operating state according to the task list. If necessary, the operating parameters corresponding to each task are determined, and the matching coolant temperature and flow rate parameters are found in the preset table. Then, the method comprehensively considers multiple influencing factors of the delivery pipeline, including the number of computers passing through, the thermal conductivity of the pipe, and the length of the pipeline, and calculates the impact score of each pipeline by assigning different weights. On this basis, the delivery pipeline score list is obtained by combining the coolant parameters and the pipeline impact score, and the optimal delivery scheme under each task is screened out. Finally, the optimized delivery pipeline list is sent to the liquid cooling equipment for execution. According to the real-time task requirements of the data center, the computer operating parameters and cooling system are dynamically adjusted to avoid the problems of over-cooling or insufficient cooling in the traditional scheme, and reduce energy waste. At the same time, the differential impact of the coolant delivery pipeline is considered. By introducing the weight of the influencing factor and the scoring mechanism, the optimal delivery scheme under various working conditions is found, which further improves the efficiency and accuracy of the liquid cooling system.

[0083] For the number of computers, the method introduces a preset threshold. When the actual number exceeds the threshold, a larger weight is given, otherwise a smaller weight is given. This way of distinguishing weights can highlight the impact of pipeline loads, give more weight to pipelines in computer-intensive areas, and ensure the rational allocation of cooling resources. At the same time, a similar threshold comparison strategy is also adopted for thermal conductivity. Different weights are given according to the actual thermal conductivity of the pipe to emphasize the impact of material selection on heat transfer efficiency. In addition, the method also stipulates that the sum of the weights of the three influencing factors must be equal to the preset value, forming a dynamic balance weight allocation mechanism. This means that when the weight of a factor increases, the weights of other factors will decrease accordingly, thereby maintaining the total weight unchanged. On the one hand, this weight allocation method ensures the stability of the impact assessment and avoids the excessive dominance of a certain factor. On the other hand, it also increases the flexibility of weight assignment, and can adjust the emphasis of different factors according to actual needs.

[0084] First, it is necessary to obtain the parameter values ​​of the influencing factors corresponding to each pipeline, and then input the parameter values ​​and weights into the impact scoring function, and the function is used to calculate the impact score of the pipeline. The impact scoring function uniformly expresses the effects of multiple influencing factors in a quantitative way, providing an operational mathematical model for the comprehensive evaluation of pipeline performance. Through function calculation, the degree of influence of different factors is converted into a unified scoring value, making the pipeline properties that were originally difficult to directly compare measurable and sortable. This quantitative evaluation method overcomes the subjectivity and ambiguity of qualitative analysis and enhances the scientificity and accuracy of pipeline selection. On the other hand, the design of the impact scoring function also provides an important basis for the optimization and control of the liquid cooling system. By analyzing the scoring results, the pros and cons of each pipeline can be clearly judged and the key factors affecting the heat transfer efficiency can be identified. This information can guide engineers to make targeted improvements in pipeline layout, material selection and other aspects, and continuously improve the performance of the liquid cooling system. At the same time, the pipeline impact score also lays the foundation for subsequent advanced strategies such as flow rate and temperature optimization and multi-pipeline collaborative control, enabling the liquid cooling system to achieve intelligent and refined regulation in a larger range.

[0085] The function comprehensively considers the three influencing factors of the number of computers in the passage, the thermal conductivity of the pipe, and the length of the pipe, and unifies their effects into a score value by weighted summation. This quantitative evaluation method overcomes the subjectivity and uncertainty of traditional qualitative analysis, and provides a reliable mathematical basis for the comparison and optimization of pipeline performance. The design of the scoring function is reasonable, concise, and has a clear physical meaning. Among them, the number of computers reflects the cooling load borne by the pipeline, the thermal conductivity represents the heat transfer performance of the pipe, and the pipe length affects the resistance loss of the fluid. By introducing the weight coefficient, the function can flexibly adjust the importance of each factor to meet the needs of different application scenarios. At the same time, the sum of the weights in the function is limited to 1, which is both a normalization process and ensures the stability and comparability of the scoring results.

[0086] The score combines the effects of factors such as coolant temperature, flow rate and pipeline impact score, and quantifies the comprehensive performance of the pipeline through a mathematical function. The pipeline score reflects the mutual influence between the pipeline and the coolant, and reflects the overall effect of the liquid cooling system. In the design of the pipeline scoring function, the temperature, flow rate and impact score appear in the form of exponential function, reciprocal and inverse respectively. The selection of this form has a clear physical meaning. First, the exponential function is a common attenuation model. As the temperature increases, its reciprocal decays exponentially, indicating the nonlinear relationship between the negative impact of temperature on the cooling effect and the temperature increase. Secondly, the flow rate and the inverse of the impact score reflect the inverse relationship between them and the heat transfer performance of the pipeline, that is, the greater the flow rate and the higher the impact score, the worse the heat transfer effect per unit length of the pipeline. By comprehensively considering the effects of temperature, flow rate and impact score, and using the characteristics of reciprocal operation, the pipeline scoring function cleverly unifies their influences into a dimensionless score value.

[0087] When the system determines that there is no need to change the operating state, it maintains the current coolant parameters and delivery pipeline selection. Once it is determined that the existing cooling configuration can meet the task requirements, the system will maintain the current operating parameters and avoid unnecessary adjustment behavior.

[0088] In the above embodiment, step S104: determining a plurality of delivery pipelines for the liquid cooling device to deliver the cooling liquid to the computer will be described through the following embodiment. Figure 2 , another data center liquid cooling method in the embodiment of the present application is described: See also Figure 2 , is another flow chart of a data center liquid cooling method in an embodiment of the present application.

[0089] S201, obtaining a pipeline layout diagram for connecting a liquid cooling device and a computer through a delivery pipeline line; In the data center liquid cooling system, the pipeline layout diagram is an important design document that intuitively shows the physical connection between the liquid cooling equipment and the computer. Through the pipeline layout diagram, the system can clearly understand the paths and nodes that the coolant passes through from the liquid cooling equipment to the heat dissipation components of each computer. This provides the necessary data basis for the subsequent identification and optimization of the delivery pipeline line.

[0090] Specifically, the system first needs to obtain a complete pipeline layout diagram, which should include the following key information: the location and parameters of the liquid cooling equipment, the location and model of the computer, the direction and connection points of the transmission pipeline, the distribution of control devices such as valves and flow meters, etc. This information can come from the CAD drawings provided by the design department, or it can be obtained through manual on-site surveys or automated mapping tools.

[0091] S202, using a path recognition algorithm to calculate the pipeline layout diagram to obtain a number of pipeline routes for the liquid cooling device to transport the coolant to the computer; After obtaining the pipeline layout diagram, the system needs to further analyze the connection relationship in the diagram and identify the pipeline routes from the liquid cooling equipment to each computer. This is actually a path search problem, which can be solved efficiently using graph theory algorithms. Through path identification, the system can obtain several feasible pipeline routes, providing a decision-making basis for subsequent path optimization and coolant scheduling.

[0092] Specifically, the system first converts the pipeline layout diagram into a directed graph model, in which the nodes represent liquid cooling equipment, computers, and pipeline connection points, the edges represent pipeline line segments, the direction of the edge represents the flow direction of the coolant, and the weight of the edge can represent properties such as pipeline length or flow resistance.

[0093] Then, the system uses the liquid cooling device as the starting point and each computer as the end point, and uses classic path search algorithms such as depth-first search (DFS), breadth-first search (BFS), Dijkstra shortest path algorithm, etc. to traverse the entire directed graph and find all paths from the starting point to the end point. These paths correspond to the pipelines that the liquid cooling device uses to transport coolant to each computer.

[0094] S203, determining the number of route computers, pipeline materials and pipeline lengths corresponding to a plurality of transmission pipeline routes according to the pipeline layout diagram.

[0095] After obtaining the pipeline route from the liquid cooling equipment to each computer, the system needs to further analyze the key attributes of each route, such as the number of computer nodes passed, pipeline material, pipeline length, etc. These attributes directly affect the flow characteristics and heat transfer efficiency of the coolant in the pipeline, and are an important basis for subsequent optimization of pipeline paths and flow regulation. Through attribute analysis, the system can comprehensively evaluate the physical characteristics of each pipeline route, laying the foundation for refined liquid cooling control.

[0096] Specifically, for each pipeline route, the system first counts the number of computer nodes it passes through. This attribute reflects the complexity of the pipeline route and the required flow distribution accuracy. The more nodes there are, the more complex the topology of the pipeline route is, and the more difficult it is to regulate the flow of the coolant. Therefore, when optimizing the pipeline path, try to choose a route with a small number of nodes to reduce control difficulty and energy loss.

[0097] Secondly, the system identifies the material information of the pipeline line. Pipes of different materials have different physical properties such as thermal conductivity, specific heat capacity, and corrosion resistance, which directly affects the temperature change and flow resistance of the coolant in the pipeline. Common liquid cooling pipeline materials include copper, stainless steel, polytetrafluoroethylene, etc. Copper pipes have good thermal conductivity, but the cost is relatively high; stainless steel pipes have high strength, but relatively low thermal conductivity; polytetrafluoroethylene pipes have good insulation, but poor pressure resistance. The system needs to select the appropriate pipeline material according to actual working conditions and design requirements, and consider the influence of the material in temperature and flow control.

[0098] Finally, the system calculates the length of each delivery pipeline line. Pipeline length is a key factor affecting coolant flow resistance and heat transfer loss. At the same flow rate, the longer the pipeline, the greater the coolant flow resistance and the higher the pump power consumption; at the same time, the longer the pipeline, the larger the heat exchange area between the coolant and the pipe wall, and the more serious the temperature loss. Therefore, when optimizing the pipeline path, you should try to choose a shorter line to reduce flow resistance and heat transfer loss.

[0099] The above embodiments have the following beneficial effects: First, it is necessary to obtain the pipeline layout diagram, and then use the path recognition algorithm to automatically calculate all possible transmission routes. This method can quickly and accurately find the effective path in the complex pipeline network inside the data center, and extract the key parameters that affect the cooling effect according to the pipeline layout diagram, including the number of computers passing through the pipeline, the properties of the pipe, and the length of the pipe. These three parameters reflect the characteristics of the transmission pipeline from different angles: the number of computers passing through affects the distribution of cooling load, the properties of the pipe determine the heat conduction efficiency, and the pipe length is related to fluid resistance and transmission loss. By collecting these parameters, the subsequent evaluation algorithm can analyze the advantages and disadvantages of each pipeline route more comprehensively and finely, providing a reliable basis for finding the best transmission solution. The automatic search and parameter extraction of the transmission line are realized, which improves the control efficiency and accuracy of the liquid cooling system.

[0100] The following describes the system in the embodiment of the present invention from the perspective of hardware processing. Figure 3 , which is a schematic diagram of the physical device structure of a data center liquid cooling system provided in an embodiment of the present application.

[0101] It should be noted that Figure 3 The structure of the system shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0102] like Figure 3As shown, the system includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 302 or the program loaded from the storage part 308 to the random access memory (R access memory, R data center liquid cooling method, system, storage medium and program product M) 303, such as executing the method in the above embodiment. In the R data center liquid cooling method, system, storage medium and program product M303, various programs and data required for system operation are also stored. CPU301, ROM302 and R data center liquid cooling method, system, storage medium and program product M303 are connected to each other through a bus 304. Input / output (I / O) interface 305 is also connected to the bus 304.

[0103] The following components are connected to the I / O interface 305: an input section 306 including a camera, an infrared sensor, etc.; an output section 307 including a liquid crystal display (LCD) and a speaker, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN card, a modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. A removable medium 311, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 310 as needed so that a computer program read therefrom is installed into the storage section 308 as needed.

[0104] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication part 309, and / or installed from a removable medium 311. When the computer program is executed by the central processing unit (CPU) 301, various functions defined in the present invention are performed.

[0105] It should be noted that the computer-readable medium shown in the embodiments of the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection with one or more conductors, a portable computer disk, a hard disk, a random access memory (R), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, apparatus, or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, wherein a computer-readable computer program is carried. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof.

[0106] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Among them, each box in the flowchart or block diagram can represent a module, a program segment, or a part of the code, and the above-mentioned module, program segment, or a part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0107] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the system described in the above embodiment; or may exist independently without being assembled into the system. The above storage medium carries one or more computer programs, and when the above one or more computer programs are executed by a processor of a system, the system implements the method provided in the above embodiment.

[0108] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0109] As used in the above embodiments, the term "when..." may be interpreted to mean "if..." or "after..." or "in response to determining..." or "in response to detecting...", depending on the context. Similarly, the phrases "upon determining..." or "if (the stated condition or event) is detected" may be interpreted to mean "if determining..." or "in response to determining..." or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)", depending on the context.

[0110] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk), etc.

[0111] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by a computer program to instruct the relevant hardware, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-mentioned method embodiments. The aforementioned storage medium includes: ROM or random access memory R data center liquid cooling method, system, storage medium and program product M, disk or optical disk and other media that can store program codes.

Claims

1. A data center liquid cooling method, characterized in that: include: Determining whether the computer needs to change its operating state according to a list of pending tasks, wherein the list of pending tasks includes pending tasks after the current time, and the computer corresponds to different operating states when executing different tasks; If so, determining the computer operating parameters corresponding to each task according to the order of tasks in the to-be-completed task list, and obtaining an operating parameter list whose arrangement order corresponds to the order of tasks one by one; Determining a coolant parameter list corresponding to the operating parameter list in a preset table, the operating parameters corresponding to the coolant parameters one by one, the coolant parameters including the coolant temperature and the coolant flow rate; determining a plurality of delivery pipeline routes for a liquid cooling device to deliver cooling liquid to the computer; Determine the weights of coolant influencing factors in the plurality of delivery pipelines respectively, wherein the coolant influencing factors include the number of computers passing through, the material of the pipeline and the length of the pipeline; Calculate the impact score corresponding to each delivery pipeline line according to the weight of the coolant impact factor; A pipeline score list is obtained based on the coolant parameters and the impact score of each pipeline line. The pipeline score list includes each of the tasks to be completed, each of the tasks to be completed includes a corresponding pipeline score, each of the tasks to be completed and the corresponding pipeline score constitute a pipeline score sublist, and each of the pipeline score sublists constitutes the pipeline score list; Determine the best delivery pipeline score with the highest score in each delivery pipeline score sublist respectively, and obtain the best delivery pipeline score list, wherein the best delivery pipeline score corresponds to the best delivery pipeline route; An optimal delivery pipeline list is generated according to the optimal delivery pipeline score list, and the optimal delivery pipeline list is sent to the liquid cooling device, and the arrangement order of the optimal delivery pipeline list corresponds to the task order of the to-be-completed task list.

2. The method according to claim 1, characterized in that The step of determining a plurality of delivery pipelines for delivering cooling liquid from a liquid cooling device to the computer specifically includes: Obtaining a pipeline layout diagram connecting a liquid cooling device and the computer through a delivery pipeline line; The pipeline layout diagram is calculated using a path recognition algorithm to obtain a number of pipeline routes for the liquid cooling device to transport the cooling liquid to the computer; The number of path computers, the material of the pipeline and the length of the pipeline corresponding to the plurality of delivery pipeline lines are determined respectively according to the pipeline layout diagram, wherein the number of path computers, the material of the pipeline and the length of the pipeline are coolant influencing factors.

3. The method according to claim 1 or 2, characterized in that: The step of respectively determining the weights of the coolant influencing factors in the plurality of delivery pipelines specifically includes: Determining whether the number of computers in the path is greater than a preset number; If it is greater than the preset number, the weight of the number of computers in the path is set to the first weight; If it is not greater than the preset number, setting the weight of the number of computers in the path to a second weight; determining an average thermal conductivity of the pipe material; Determining whether the average thermal conductivity is greater than a preset thermal conductivity; If it is greater than the preset thermal conductivity, the weight of the pipeline material is set to a third weight; If it is not greater than the preset thermal conductivity, setting the weight of the pipeline material to a fourth weight; The weight of the pipeline length is set as the fifth weight, and the sum of the weight of the number of computers in the path, the weight of the pipeline material and the weight of the pipeline length is equal to a preset threshold.

4. The method according to claim 1, characterized in that: The influence score corresponding to each delivery pipeline line is obtained by calculating the weight of the coolant influence factor, specifically including: Determine the parameter value of the coolant influencing factor corresponding to each delivery pipeline line; The parameter values ​​corresponding to each transmission pipeline route and the weight of the coolant influencing factor are respectively input into the influence scoring function to obtain the influence score corresponding to each transmission pipeline route.

5. The method according to claim 4, characterized in that The influence scoring function is: In the formula, Score the impact, is the number of computers in the path, is the first weight or the second weight, is the average thermal conductivity, is the third weight or the fourth weight, is the pipeline length, is the fifth weight.

6. The method according to claim 1, characterized in that The pipeline score list is obtained by calculating the coolant parameters and the impact score of each pipeline line, specifically including: Inputting the coolant parameters and the impact score of each delivery pipeline route into a delivery pipeline score calculation function to obtain a delivery pipeline score for each delivery pipeline route; Add the delivery pipeline score corresponding to the coolant parameter to the corresponding blank list to obtain a delivery pipeline score list; The pipeline scoring calculation function is: In the formula, Score the pipeline, is the coolant temperature, is the coolant flow rate, rate the impact; In the and stated When the coolant temperature is a fixed value, the higher the coolant temperature, the smaller the effect of the delivery pipeline on the coolant. The bigger; In the and stated When the coolant flow rate is a fixed value, the greater the coolant flow rate, the greater the impact of the delivery pipeline on the coolant. The smaller; In the and stated When the impact score is a fixed value, the greater the impact score, the greater the impact of the delivery pipeline on the coolant. The smaller.

7. The method according to claim 1, characterized in that After generating an optimal delivery pipeline list according to the optimal delivery pipeline score list and sending the optimal delivery pipeline list to the liquid cooling device, the method further includes: If not, the coolant parameters and the delivery pipeline at the current moment are maintained to cool the computer.

8. A data center liquid cooling system, characterized in that: The system comprises: One or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the system to execute the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instructions are executed on a system, the system is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on a system, the system is caused to execute the method according to any one of claims 1 to 7.

Images